-
-
Notifications
You must be signed in to change notification settings - Fork 29k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Beta 0.114 All OZW devices report unavailable untill used #38658
Comments
This sounds like symptoms from the cache invalidation that was part of the zwave add-on version 0.5.2. |
ozw documentation |
The ozw cache file has been rebuild sometime ago if that is what you mean? |
I did some more testing and when I reconfigure mqtt from the integrations menu all devices are immediately available again. So this is not ozw related, but related to mqtt. |
mqtt documentation |
Hey there @home-assistant/core, @emontnemery, mind taking a look at this issue as its been labeled with an integration ( |
That PR can't be related, it only changed some tests. What are the commit IDs for 0.114.0-dev20200724 and 0.114.0-dev20200723? |
Those are the nightly dev builds which I pulled from docker hub. |
There doesn't seem to be any changes to the mqtt integration in that date range. Would you be able to capture some log info during a start of home assistant with both the "bad" and "good" build? logger:
default: info
logs:
homeassistant.core: debug
homeassistant.components.mqtt: debug
homeassistant.components.ozw: debug If you can capture the docker ozwdaemon logs, that could also help. |
Note: I edited @MartinHjelmare to request log both from the good and bad version. |
I attached both the log files. I used 0.113.3 and 0.114b4, hopefully that is ok too. OZW logs can be done, but since I'm running two installations that would take some more work. |
Assuming it's a problem for all ozw entities, here's an example:
Failing log: MQTT message for
|
@VinceRMP In "bad" case log, it seems not all OZW devices are stuck as |
It happens to all ozw devices unless they receive an update in case of an sensor or I manually switch the device on or off. When I reboot HA some of the devices will return as available but this happens because they are triggered for example the motion sensors. Also it seems an automation is able to activate an switch, but I can't do it from the UI. I haven't checked it fully, but I noticed this earlier today. |
The ones I notice that it does not work is for example: Most switches also fail, for example These device are always turned on so don't change the status. |
For switch.synology (working case):
For switch.synology (failing case):
|
|
Is the 30 seconds a problem? Do you have an idea why on the latest releases de states are not updated? Could it be a timing issue perhaps? |
ozw documentation |
@MartinHjelmare can you take a look again? From what I can understand, MQTT messages are received and acknowledged by the ozw integration. |
Hey there @cgarwood, @marcelveldt, mind taking a look at this issue as its been labeled with an integration ( |
I just checked it and that solves the problem it seems. Did a few reboots and ozw works everytime after the reboot. Tried it on the latest 0.115 dev build. If you want I can also try it on the 0.114b4 build? |
I think the problem is that we create the entity too early before the complete node state is reported from the broker. The polling worked around that issue by scheduling a state update after 30, or so, seconds. When we removed the polling, the issue surfaced. |
We've been looking at this, but fail to reproduce it. The polling lead seems to be a false lead. The ideas I had seem to be dead ends. Two different test environments are working good with latest dev branch. Can we get an mqtt dump where the problem exists? |
Yes sure no problem. Just with the default settings and directly after the reboot I assume? |
Yes, that sounds good. Increase the listening time to at least 30 seconds to capture all the messages. |
See the attached file, this was run on the latest nightly dev build. (If this really only happens on my side, I can automate it now to solve the issue with the service call) |
Can you test with the Supervisor zwave add-on for the ozwdaemon instead of standalone docker? It would be interesting to see if there's a difference in behavior. |
I could but that would need some reconfiguration. What you want me to do is create a new system using Home Assitant OS, load my current configuration. |
Yes, if you can do that, it could help pinpoint the issue. I understand that's not very easily done. If you don't have a spare computer it might be too much work. |
It's no problem. I'm glad I can help with the issue. |
If it still gives problems, do you want some log files? |
Thanks for your assistance on this. We were not able to reproduce the issue on our environments so would be great if you can help us out testing. We want to make sure there isn't some global issue introduced in recent changes. |
For now the test with the OZW daemon addon will suffice. After that we can do some more debugging with logs. |
No problem, I'm a happy HA user and since this is the only way I can contribute I'm happy to do so. |
Ok so I have some good news I guess. I can't reproduce it either when using the asked configuration. Just to make sure I have activated most of the battery devices, only missing 7 which aren't easy to reach. So it doesn't seem to be a global issue. Still not sure what the problem on my side is though.. and if other people will have a problem when using external docker applications. |
Big thanks for taking time to test so thoroughly! We can look at the changes in OpenZWave and ozw daemon that have happened in the versions used in the add-on compared to the docker image you're running. Maybe there are clues there? |
No problem, glad I can help. Edit: I just realised that my last test with my production system does not have to be correct.. Earlier today I made an automation which executes the mqtt.dump, which force the refresh needed to make it work. I could do that test again if needed to make sure, but that won't be today. |
If the
|
I checked the functionality again from my production system (0.114.0b4) against the HA mqtt test connected to the seperate OZW container. This does work with the mqtt dump automation disabled. So changing mqtt server causes the problem somehow. Attached two mqtt dumps from the mosquitto_sub command (30 seconds) |
The problem
When I reboot homeassistant all z-wave devices are unavailable and unusable untill I manually use the device or it reports some information in case it is a sensor.
So for a switch I have to manually turn device on or off to make it available again.
Sensors become available again when they send an update.
When I force a reboot of the (external) OZW container everything is available when OZW has finished loading. I noticed this on the dev builds too, but since that is development that can happen. I tried the several dev builds again to see when it stopped working for me and that is 0.114.0-dev20200724. So 0.114.0-dev20200723 does work as expected. My production system (Home Assistant OS) which runs 113.3 kept working without any problem and keeps receiving all the information.
After updating this to the latest beta the same problem occured as on my test system (Home Assistant Container).
I have around 81 devices z-waves in my network.
Environment
Problem-relevant
configuration.yaml
Traceback/Error logs
No errors reported
Additional information
The text was updated successfully, but these errors were encountered: